Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    INDEX
    Explanations

    the vocal filler "um" used in speech

    oai_token-act-pair · gpt-3.5-turbo

    fillers or hesitation sounds in speech

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    hips
    -0.91
    Kings
    -0.84
    lines
    -0.75
    bedroom
    -0.71
    UGC
    -0.70
    ingham
    -0.69
    BOOK
    -0.68
    Monitor
    -0.68
    Pred
    -0.68
    HEAD
    -0.68
    POSITIVE LOGITS
     um
    1.09
    Um
    0.93
     awa
    0.84
    ugh
    0.80
    pport
    0.80
    ufact
    0.80
     Um
    0.79
     passer
    0.74
    uge
    0.72
     fu
    0.72
    Act Density 0.006%

    No Known Activations